Where to Begin
The shift to a fully market-based electricity pricing system has significantly slowed solar power investment over the past year. This marks a sharp contrast to the previous five years, which saw unprecedented expansion in China's solar photovoltaic (PV) sector.
From 2020 to 2025, the country's annual新增 photovoltaic capacity surged from 48.2 GW to 317 GW, a roughly 6.6-fold increase, while cumulative installed capacity rose over fourfold to 1,200 GW. State-owned power enterprises and local government companies channeled funds into large-scale, centralized solar plants in the northwest and northeast. In central and eastern regions, private developers and equipment manufacturers expanded into factories, logistics parks, and rural rooftops, driving the growth of commercial, industrial, and residential distributed solar.
Data from the National Energy Administration shows that in the first half of 2026, China added 72.07 GW of new solar capacity, a 66% year-on-year decline. This included 29.55 GW of new centralized plants and 42.22 GW of distributed projects, down 70% and 60% respectively from the same period in 2025.
A pivotal policy shift came on June 1, 2025, with the implementation of the "Notice on Deepening the Market-Oriented Reform of New Energy On-Grid Prices to Promote High-Quality Development" (known in the industry as "Document No. 136"). Issued by the National Development and Reform Commission and the National Energy Administration, this regulation directs that the vast majority of electricity from newly built solar and wind projects must now be priced through market competition. Previously, investors could lock in construction costs, operational hours, and financing rates to calculate a relatively stable internal rate of return before breaking ground, effectively allowing them to "earn money without effort." Now, earnings are no longer guaranteed by fixed tariffs and are instead subject to real-time supply and demand.
Faced with uncertain market-based electricity prices, central and state-owned enterprises have slowed their investments. Smaller-scale plant owners are abandoning the incremental market, pivoting instead to projects with higher rates of self-consumption to minimize grid supply. They are also exploring new revenue streams, such as green electricity trading, virtual power plants, and zero-carbon industrial parks.
Fixed Tariffs Fade Out, Solar Revenue Gets Repriced
The revenue mechanism for solar plants has undergone multiple rounds of reform, gradually shifting from fixed tariffs to prices set by market supply and demand. Since 2021, central government subsidies have been completely phased out for newly registered centralized solar plants and commercial-industrial distributed projects. For these, on-grid prices are generally pegged to the local benchmark coal-fired power price, though voluntary market participation is allowed. For projects selling all their power to the grid, the coal benchmark price provides a relatively stable income. For those using a "self-consumption, surplus-on-grid" model, revenue comes from savings on commercial electricity bills and the sale of surplus power to the grid. Investors could previously calculate earnings based on generation hours, with the electricity price being a relatively fixed variable.
In February 2025, "Document No. 136" mandated that new energy power generation should, in principle, enter the electricity market entirely, with prices formed by the market, and established a sustainable pricing mechanism for new energy. This means solar plants, once reliant on stable on-grid prices, now face market price volatility. For existing projects commissioned before June 1, 2025, the mechanism price follows original policy but must not exceed the local coal benchmark price. For projects commissioned after that date, provincial governments determine the scale of mechanism electricity and hold auctions to determine a "floor price." Solar projects must first sell their power at market prices. If the average market price is lower than the mechanism price, they receive a compensation payment for the difference; if it's higher, they must pay back the excess.
As of July 2026, all Chinese provinces have completed their first round of mechanism price auctions covering projects commissioned between June 1, 2025, and December 31, 2026, with four provinces having completed auctions for the first half of 2027. Bloomberg NEF estimates that the total winning solar capacity for the 2025-2026 period was about 101 GW, representing only 20% of all commissioned solar projects. Projects not covered by the mechanism electricity quota bear the full brunt of market price fluctuations.
The impact is felt daily. From dawn, as solar generation ramps up, spot prices begin to fall, hitting their lowest point during the sunniest midday hours. Massive solar output depresses real-time prices, especially in areas with concentrated solar installations, limited local load, and weak transmission capacity. This leads to faster and more severe declines in the actual settlement price for solar projects. BNEF estimates that the actual market price for solar power is, on average, 127 yuan/MWh (0.127 yuan/kWh) lower than the mechanism price.
Auction results for mechanism prices vary significantly by region. Shanghai's 2025 mechanism price for new projects was 0.4155 yuan/kWh, while Gansu province's two auction rounds both resulted in 0.1954 yuan/kWh—a more than twofold difference. While project structures and auction rules differ, this price gap highlights that local load, grid absorption capacity, and new energy supply conditions are now critical determinants of a plant's value.
BNEF categorizes provincial markets into four types: demand-driven markets like Shanghai and Jiangsu, with strong local absorption; areas rich in hydropower like Sichuan and Yunnan, heavily influenced by wet and dry seasons; regions in rapid expansion and rule adjustment like Shandong, Hebei, and Shanxi; and resource-rich areas like Gansu, Xinjiang, Qinghai, and Ningxia, vulnerable to insufficient local load and transmission constraints.
With the full marketization of new energy, investors must not only secure mechanism prices but also find stable demand for green electricity and long-term power purchase agreements. To address this, policy focus has shifted to promoting absorption. In May 2026, the "Notice on the Orderly Promotion of Direct Green Electricity Connections for Multiple Users" was issued, allowing multiple consumers to jointly connect to nearby new energy projects, specifying conditions for implementation in industrial parks, data centers, and export-oriented manufacturers. In June, the "Implementation Measures for Minimum Renewable Energy Consumption Targets and Renewable Energy Power Consumption Responsibility Weights" were introduced, requiring regions and key energy-consuming industries to meet a certain proportion of renewable energy consumption, offering a new pathway for solar projects to find long-term buyers.
Centralized Plants: State-Owned Enterprises Turn Cautious
Centralized solar plants, typically built on a large scale on ground or water surfaces, rely primarily on selling power to the grid for revenue. With individual project capacities ranging from 100 MW to over 1 GW, they have longer construction periods and larger investment requirements compared to distributed projects. State-owned enterprises (SOEs) are the primary investors. Facing revenue uncertainty, they have slowed their investments and are accelerating the disposal of stalled pre-approved projects.
The income of a centralized plant is directly dependent on the market price and mechanism price in its province. In addition to price volatility, absorption bottlenecks create uncertainty. In the first quarter of 2026, the national solar utilization rate fell to 91.2%, dropping below 80% in northwestern regions like Gansu, Qinghai, and Xinjiang. These areas have limited local electricity load but are home to most new renewable capacity, and power absorption is constrained by transmission capacity. Even after construction, plants may not be able to sell their predicted generation to the grid.
The shift in SOE investment reflects this pressure. Industry media SolarPower reports that the average growth rate of new energy investment by central SOEs in 2025 fell to 22%, down from 32% in 2024. Specifically, solar investment growth dropped from 57% to 29%, a 28-percentage-point decline. According to an incomplete tally by Polaris Solar PV Network, since 2026, SOEs have listed 62 new energy companies' equity for transfer, with a combined base price exceeding 2 billion yuan, including over 30 entities selling all their shares.
In 2026, local governments began clearing out solar projects that had long occupied construction quotas and grid connection resources. In January, Shandong province removed 63 centralized solar projects with a total capacity of 5.839 GW from its list for failing to connect to the grid on schedule, while granting a grace period until the end of 2026 for 29 projects that had started construction. In June, Fujian province scrapped 10 centralized "fishery-solar complementary" projects totaling 958.8 MW. In July, Guizhou province cleaned up 11 projects totaling 1.09 GW that had not completed approval on time, had not started construction, or had been voluntarily abandoned.
SOEs are also setting their own investment return thresholds. In December 2025, Changyuan Electric Power, a subsidiary of China Energy Investment Corporation, reduced three solar projects in Hubei province by a total of 800 MW, reducing investment by about 4.694 billion yuan. According to Changyuan Electric's recently revised investment management rules, the internal rate of return on equity for new solar and wind projects must be at least 6.5%, and the pre-tax internal rate of return on total investment must be at least 6%. Before "Document No. 136," the internal rate of return for centralized solar projects could reach 7% or even approach 10%.
Policy signals reinforce this trend. Centralized solar will continue to expand, but new projects must simultaneously address power transmission and stable absorption. The "Renewable Energy Development 15th Five-Year Plan" proposes adding over 370 GW of new wind and solar capacity in the "Three Norths" regions during the plan period, and continuing to plan and reserve external transmission bases in Inner Mongolia, Gansu, Qinghai, Ningxia, and Xinjiang. The plan also requires that by 2030, the "confidence output" of new centralized wind and solar stations must be at least 10%, with better-performing projects encouraged to exceed 20%.
Distributed Plants: Small and Medium Enterprises Restructure Revenue Models
Distributed solar, built on residential rooftops, commercial factory roofs, and industrial park roofs, features smaller individual scales and a large number of projects, with development heavily reliant on local resources. Some investors are small and medium-sized enterprises who are familiar with local conditions but have limited risk tolerance. After "Document No. 136," the distributed solar sector underwent a shakeout. Many companies have exited the market entirely, while those still investing are more cautious with new projects, pivoting to those with a high proportion of self-consumption and seeking new revenue sources through green electricity trading, solar-storage integration, and virtual power plants.
Before the full marketization of new energy, "full-sale" distributed solar projects sold all their power to the grid at a relatively stable on-grid price. The revenue for "self-consumption, surplus-on-grid" projects had two parts: electricity used directly by the consumer replaced grid-purchased commercial power, with the benefit being the difference between the commercial electricity price and the solar generation cost; and the surplus power was sold at the on-grid price. Construction costs, generation hours, and electricity prices were all relatively fixed variables, allowing investors to calculate project returns easily.
Now, the surplus-on-grid portion faces market price volatility, making the full-sale model largely unviable. At the same time, many provinces have adjusted the midday period, when solar output is highest, to be an off-peak time. This lowers the price of grid power during the day, thereby reducing the savings from self-consumption. According to a report by the Rocky Mountain Institute, as of November 2025, 24 provinces had adjusted parts of the midday period to off-peak. Future project development will increasingly revolve around stable loads, with investors typically requiring a self-consumption rate of 80% to 90%.
This change has made the previously common Energy Management Contract (EMC) model increasingly difficult to make profitable. Under this model, the developer invests in and owns the plant, while the rooftop user (the power consumer) avoids upfront costs by purchasing power at a discounted price. The developer recovers its investment through long-term electricity sales revenue. Now, with the consumer's power purchase price becoming unstable, it's difficult to calculate the discount rate, the savings allocation, and the developer's revenue over a multi-year contract using past models.
Shandong province, which has the largest installed solar capacity and was an early adopter of spot market trading, illustrates this shift most clearly. Shandong has long been the leading province in solar capacity and is also a high-load electricity consumer. In 2025, Shandong is expected to import over 160 TWh of electricity, about 20% of its total social electricity consumption. In 2025, the first national auction for new energy mechanism prices was held in Shandong, resulting in a solar mechanism price of 0.225 yuan/kWh, which rose to 0.261 yuan/kWh in 2026—still below the local coal benchmark price of 0.395 yuan/kWh, placing it in a "price trough."
Furthermore, Shandong has clarified that from 2027 onwards, non-natural person household distributed solar will no longer be included in the mechanism price auction. This means only solar projects self-invested by individual farmers can enjoy the mechanism price. Contacted by Caijing, one industry source noted that this policy has severely impacted the profitability of distributed solar businesses, making household projects nearly unattractive. Yujin New Energy, a company investing in solar projects in Shandong, has therefore completely abandoned the household leasing model. It is winding down its existing portfolio and has shifted its new household business to offering solar loans to individual users. It has already implemented over 2,000 solar loan projects in Shandong, with a loan scale approaching 200 million yuan.
According to Yujin New Energy Chairman Song Changfang, with the mechanism price providing a safety net, a 30 kW household solar system in Shandong can still generate an annual income of about 10,000 yuan for the farmer. Under the solar loan model, the farmer buys and owns the plant through a loan, while the company earns revenue from equipment sales and subsequent operations and maintenance, ensuring the long-term stable operation of the generation equipment and providing the user with stable returns. The company's commercial and industrial business has pivoted to seeking明确的 electricity consumption scenarios. Song Changfang told Caijing that Yujin New Energy's commercial projects are now primarily focused on solar-storage integration, prioritizing clients with stable power consumption like packaging factories, commercial centers, and industrial properties, and establishing partnerships with export-oriented companies needing green certificates. Additionally, the company is developing heavy-truck charging stations as a new consumption scenario, planning solar carports and storage, aiming for about 20% of the station's power to come from solar, and is preparing to establish a virtual power plant based on its electricity sales license.
Zhang Xiaobin, Vice President of the Shandong Renewable Energy Industry Association, believes that stable power sources like coal, nuclear, and self-owned power plants in Shandong cannot significantly reduce their output, and there is insufficient regulation from pumped storage hydro. When load is low, the only option is to curtail wind and solar output, leading to significant revenue losses. "Building distributed solar used to rely mainly on local resources. Now, you need capabilities in power market management and load management. The problems Shandong's solar industry is facing now will definitely appear in other provinces in the future," he told Caijing.
Compared to Shandong, the Jiangsu solar market benefits from its strong local load and commercial-industrial absorption capacity. With its dense manufacturing base and abundant factory rooftop resources, solar development in Jiangsu is primarily distributed. By the end of 2025, the province's total distributed solar installed capacity reached 64.01 GW, ranking first in the country. Many distributed solar projects are built on factory rooftops, and the daytime production load aligns well with solar output, leading to a relatively high absorption rate. Jiangsu's "15th Five-Year Plan" targets adding another 56 GW of distributed solar capacity by 2030.
Several industry insiders at Jiangsu-based solar investment companies told Caijing that around 2021, driven by high return expectations, project development prioritized securing rooftop scale, with many capital investors lacking experience in power markets entering the solar sector, leading to irrational investment decisions. The full marketization of new energy will temporarily curb investment, but it will also help filter out companies with stronger operational capabilities. Currently, investors and financial leasing institutions are re-evaluating revenue models, with almost all companies demanding that the majority of power from new projects be self-consumed.
In this context, green electricity trading has become a crucial channel for solar projects to obtain market-based revenue. As of July 2026, the Jiangsu Power Exchange Center has organized 65 monthly and intra-month green electricity trading sessions, a 47.8% increase in trading frequency compared to the full year of 2025. The head of power trading at solar plant service company Linyi Zhineng told Caijing that Jiangsu has many export-oriented companies with strong demand for green electricity and green certificates, and the market supply is relatively tight, providing an additional revenue source for distributed solar projects. However, companies need to purchase green electricity from dozens or even hundreds of distributed projects, and verifying the large number of green certificates from diverse sources is difficult, limiting the acceptance of this type of certificate by power sales companies.
Aggregating distributed power through virtual power plants can also generate revenue. After Jiangsu's solar surplus power entered the market directly, the average monthly settlement price was typically around 0.22-0.23 yuan/kWh. By aggregating multiple projects to participate in green electricity trading, aggregators can sell at around 0.38 yuan/kWh. The aggregator buys power from the project owners, earning the margin while bearing the cost of generation deviation penalties and market price fluctuations.
Before the 2026 mechanism electricity declaration, Suzhou local state-owned company Suzhou Zhongxin New Energy decided to give up most of its mechanism price quota for a 50 MW wind power project in Yangzhou, choosing instead to retain the environmental attributes for participation in the green electricity market. This ensured the green power supply for its users, and the project's subsequent revenue exceeded initial expectations. Cai Jianjun, Chairman and General Manager of Zhongxin New Energy, believes that the demand for green electricity from eastern manufacturing is strong, while western regions are rich in new energy resources. To match supply and demand, in addition to building transmission channels, cross-regional green electricity trading is needed. Companies cannot view marketization solely as a risk; actively participating in green electricity trading and virtual power plants is the way to build long-term competitive advantages.
Regarding the aggregated trading of distributed solar, Cai told Caijing that current gains mainly come from three aspects: converting small-scale power into large-scale transactions; reducing generation forecast deviations through aggregation; and in some provinces, optimizing based on price differences between different projects or nodes. However, Jiangsu's relevant rules are not yet detailed, and the price optimization benefits from aggregation are not yet significant. Companies still need to assess whether the cost of building and upgrading platforms can be covered by the aggregation gains. Gu Xincen, Climate and Energy Project Officer at Greenpeace, said that for a manufacturing and export-oriented economic powerhouse like Jiangsu, the rules and pathways for distributed solar to participate in green electricity trading should be refined beyond the support of the mechanism price. Furthermore, more detailed rules for distributed solar aggregation trading are needed in the future, especially for household and small-scale commercial-industrial projects, to establish uniform rules for metering, settlement, aggregation relationships, and grid coordination.